Food Bioscience 52 (2023) 102363

Contents lists available at ScienceDirect

Food Bioscience
journal homepage: www.elsevier.com/locate/fbio

Synthesis and characterization of medium- and long-chain structural lipid

rich in α-linolenic acid and lauric acid
Ying Huang a, c, d, Wangxin Liu a, Xianliang Luo a, Minjie Zhao a, Tao Liu a, Fengqin Feng a, b, *
a
College of Biosystems Engineering and Food Science, Zhejiang University, Hangzhou, 310058, China
b
ZhongYuan Institute, Zhejiang University, Zhengzhou, 450001, China
c
College of Food Science and Engineering, Tarim University, Alar, Xinjiang, 843300, China
d
Production & Construction Group Key Laboratory of Special Agricultural Products Further Processing in Southern Xinjiang, Alar, Xinjiang, 843300, China

A R T I C L E I N F O A B S T R A C T

Keywords: Medium- and long-chain triglycerides (MLCT) have been widely concerned. In this study, lauric acid (La) was
Structural lipids incorporated into flaxseed (Linum usitatissimum L.) oil (FO, aLna-riched) by lipase-catalyzed acidolysis. The
Medium- and long-chain triglycerides synthesis efficiency of four immobilized lipases (Lipozyme RM IM, TL IM, AOAB8 and Novozym 435) was
Physicochemical properties
compared, while Lipozyme AOAB8 reaction for 2 h was the best scheme. Further, the optimal conditions for
Lauric acid
α-Linolenic acid
synthesizing structural lipids (aLSLs, rich in aLna and La) were obtained by response surface methodology: the
reaction temperature was 44.6 ◦ C, substrate molar ratio was 12 (La/FO), and enzyme loading was 9.11%. After
purification by molecular distillation, the triglycerides (TAGs) of aLSLs contained 75.35% of TAGs with
α-linolenic acid (aLna), 40.82% of MLCTs with La, and 26.75% (65.53% of the total MLCTs) of TAGs with both
aLna and La (ɑ-MLCT). Meanwhile, the content of La and aLna in aLSLs were 22.38% and 39.78%, respectively.
The chemical structure and microstructure of aLSLs were similar to FO. In addition, the thermo-oxidative
decomposition behavior of aLSLs was significantly improved compared with physical mixture oil. This
improvement might be related to the insertion of La and the structure of TAG. In all, this study firstly developed a
novel thermo-oxidatively stable MLCT with high levels of aLna and La.

1. Introduction cardiovascular disease (Zhang et al., 2010), insulin resistance (Terada

et al., 2012, 2015) and inflammation (Yue et al., 2020). As the favorable
Lipid is a vital component of the diet, which provides essential en­ health benefits of MLCT have been widely proven and accepted, the
ergy and nutrients (Guo et al., 2020; Wang et al., 2022). However, the development and efficient synthesis of novel MLCT have become the
improper intake of lipids is associated with the development of meta­ direction of researchers’ efforts to satisfy the urgent consumer demand
bolic diseases such as obesity, diabetes, and cardiovascular diseases for natural fat substitutes. However, there is a limited range of fatty
(Guo et al., 2018; Vors et al., 2020). These defects can be partially acids available in typical MLCTs on the market, with the choice of fatty
compensated by structural lipids (SLs). SLs are a kind of lipid that are acids being mainly common cooking oils combined with caprylic acid
artificially modified fatty acid composition or/and positional distribu­ (C8) and capric acid (C10) (Wang et al., 2022), leaving a gap in MLCTs
tion on the glycerol skeleton, which has better nutritional value or rich in lauric acid and α-linolenic acid.
processing properties than natural lipids (Chang et al., 2021; Guo et al., α-Linolenic acid (aLna) is an essential fatty acid, which is the raw
2020; Ji et al., 2021). Medium-chain and long-chain triglycerides material for the synthesis of docosahexaenoic acid (DHA), docosa­
(MLCTs) are the most concerned SLs, which contain both medium-chain pentaenoic acid (DPA), and eicosapentaenoic acid (EPA) in the body.
fatty acids (C8-C12, MCFA) and long-chain fatty acids (C14-C22, LCFA) in Flaxseed (Linum usitatissimum L.) oil (FO) is an ideal source of aLna
a glycerol molecule. At present, MLCT has an important contribution to (containing 30%–50%), which is widely consumed in Canada and China
make in the prevention or treatment of metabolic diseases caused by (Farag et al., 2021; Yang et al., 2021). Due to the high content of aLna,
excessive consumption of natural oils, such as obesity (Lee et al., 2022), the great health benefits of FO for cardiovascular disease, diabetes, and

* Corresponding author. College of Biosystems Engineering and Food Science, ZhongYuan Institute, Zhejiang University, No.866, Yuhangtang Road, Hangzhou,
310058, China.
E-mail address: fengfq@zju.edu.cn (F. Feng).

https://doi.org/10.1016/j.fbio.2023.102363
Received 28 October 2022; Received in revised form 3 January 2023; Accepted 3 January 2023
Available online 6 January 2023
2212-4292/© 2023 Published by Elsevier Ltd.
Y. Huang et al. Food Bioscience 52 (2023) 102363

Table 1 hydrophobic resin in the laboratory (Cai et al., 2015), optimum pH

Factors and levels for RSM. 7–10. La was obtained from Sinopharm Chemical Reagent Co., Ltd.
Level Factor (Shanghai, China). Triglyceride laurate (GTL) was purchased from Tixiai
Chemical Industry Development Co., Ltd. (Shanghai, China). Silicic acid
reaction temperature (◦ C) molar ratio (La/FO) enzyme load (%)
thin-layer chromatography (TLC) plates (10 cm × 10 cm) were sourced
− 1 40 8 6 from Shanghai Shangbang Co., Ltd. (Shanghai, China). aLna and 37 fatty
0 50 10 8
1 60 12 10
acid methyl esters standards were bought from Sigma-Aldrich Co. (St.
Louis, MO, USA). All other reagents were analytical or chromatographic
grade.
inflammation are proven (Punia et al., 2019; Wang et al., 2021; Yang
et al., 2021). Nonetheless, the oxidation sensitivity of FO hinders its 2.2. Synthesis and process optimization of aLSLs
application. Therefore, it is necessary to develop an effective method to
slow down the oxidation rate of FO (Yang et al., 2021). In this regard, 2.2.1. Lipase and reaction time screening
the introduction of oxidatively stable fatty acids into FO may be an ideal To filter the most suitable immobilized lipase and reaction time for
approach. Lauric acid (La) is a saturated MCFA containing 12 carbon catalytic acidolysis, La and FO (La/FO molar ratio 8:1) were mixed into
atoms and has good oxidation stability (Azizi et al., 2018; Panpipat rubber-capped single-necked round-bottom flasks (50 mL), and incu­
et al., 2018), and has a higher smoke point and a wider range of appli­ bated in an oscillating water bath (55 ◦ C, 200 rpm) (Jingda THZ-82;
cations compared with traditional caprylic acid and capric acid Changzhou Jintan Jingda Instrument Manufacturing Co., Ltd, Jintan,
(McCarty & DiNicolantonio, 2016). Meanwhile, its antibacterial activ­ Jiangsu, China). After homogenization the substrate mix, immobilized
ity, anti-diabetic and anti-inflammatory effects, and efficacy in reducing lipases (RM IM, TL IM, Novozym 435 and AOAB8) were added (8%, w/
cardiovascular disease were also revealed (Luo et al., 2022). Therefore, w) and reacted for 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, 6.0, and 12.0 h,
La may be a good donor of fatty acid insertions that take into account the respectively. Determination of La and aLna was described in section 2.3.
stability and health benefits of FO. All reactions were done in triplicate.
In industry, chemical or enzymatic methods are commonly used to
synthesize MLCT. Compared with chemical catalysis, enzymatic modi­ 2.2.2. Single-factor experiment
fication processes have the advantages of mild operating conditions, To get the optimal synthesis conditions of aLSLs, the following cat­
environmental friendliness, and fewer by-products (Bornscheuer, 2018). alytic factors were investigated: enzyme load (4%, 6%, 8%, 10%, and
Therefore, enzymatic reactions are widely used in the synthesis of SL, 12%, w/w), reaction temperature (40, 50, 60, 70, and 80 ◦ C), and
among which, acidolysis and transesterification enzymatic reactions are substrate molar ratio (La/FO, 4, 6, 8, 10, and 12). The main influencing
the most common. Zou et al. (2020) synthesize DHA-rich MLCT from factors and levels were obtained based on the single-factor experimental
microbial oil (from Schizochytrium sp.) and caprylic acid with Lipozyme results and applied to response surface methodology (RSM).
RM IM. Similarly, Sneha and Jeyarani (2018) obtain MLCT by incor­
porating capric acid into mango kernel fat with Lipozyme IM (1,3-sn 2.2.3. Response surface methodology (RSM)
specific, Mucor miehei). In our group previous work (Li et al., 2021), To investigate the relationship between the synthesis conditions
algal oil and La are employed to synthesize MLCT with high levels of La (reaction temperature, substrate molar ratio, enzyme load) and reaction
and DHA by enzymatic acidolysis. Based on previous studies, enzymatic degree (La inserted content and aLna cutting amount), a three-level,
acidolysis transesterification is a reliable method for synthesizing MLCT. three-factor Box-Behnken Design (BBD) (software Design-Expert
However, enzymatic reactions are sensitive to environmental factors V8.0.6.1, Stat-Ease Inc., Minneapolis, MN, USA) was implemented
such as temperature, enzyme loading, and substrates, so it is necessary (Table 1). The response variables are regression analyzed and fitted to
to optimize the reaction parameters. the following second-order polynomial model equation (Equation (1)),
In this study, MLCT rich in aLna and La (named aLSLs) was prepared which can describe the relationship between the response and the in­
by enzymatic acidolysis, using FO and La as raw materials. Firstly, lipase dependent variables.
types were selected according to enzyme activity and specificity, and the
∑
κ ∑
κ ∑κ− 1 ∑
κ
effects of different reaction conditions (reaction temperature, substrate Y = β0 + β i Xi + βii X2ii + βij Xi Xj (1)
molar ratio, enzyme load) on acidolysis reaction synthesis of aLSLs were i=1 i=1
i
j
studied to obtain better synthesis conditions. In addition, the aLSLs
obtained under the optimized conditions were purified and character­ here, Y was the variable of response; Xi and Xj were the independent
ized by UPLC-MS/MS. Finally, the lipid chemical structure and oxida­ variables; β0, βi, βii, and βij were the constant, linear coefficient,
tion thermal stability of aLSLs were evaluated. This study was aimed to quadratic coefficient, and cross-product coefficients.
provide a reference for the development of healthy and stable MLCT,
which rich in unsaturated fatty acids and La. 2.2.4. Model verification
Three times of enzymatic acidolysis were carried out under optimal
2. Materials and methods conditions to verify the credibility of the models.

2.1. Materials 2.3. GC-MS/SIM analysis of La and aLna

FO was purchased from Xinjiang Fengyi Agricultural Tourism The content of La and aLna were analyzed by GC-MS/SIM. Specif­
Development Co., Ltd. (Xinjiang, China). Lipozyme RM (lipase from ically, La and aLna standards were identified by Agilent GC7890B-
Rhizomucor miehei and fixed on resin, optimum pH 7–10), Lipozyme TL MS7000D system (GC-MS, Agilent Technologies, Santa Clara, CA,
IM (lipase from Thermomyces lanuginosus and immobilized on non- USA) combined with the Agilent HP-5MS column (30.0 m × 0.25 mm ×
compressible silica carriers, optimum pH 6–8) and Novozym 435 0.25 μm), and He (purity >99.999%) was used as carrier gas and a flow
(lipase from Candida Antarctica B and anchored on a hydrophobic car­ rate at 1 mL/min. The Agilent ALS7693-automatic sampler introduces 1
rier, optimum pH 5–9) were obtained from Novozymes Biopharma DK μL of samples and standards solution at 280 ◦ C in 100:1 split mode. Both
A/S (Bagsvaerd, Denmark). Lipozyme AOAB8 (Aspergillus oryzae lipase transfer line and ion source temperatures were set at 280 ◦ C. Charac­
powder) was bought from Levijin Bioengineering Co.td. (Shenzhen, teristic fragment ions of La and aLna were determined in m/z 50 to m/z
Guangdong, China) and immobilized with the polystyrene-based 450 in GC/MS full-scan mode with the standard substance. The

2
Y. Huang et al. Food Bioscience 52 (2023) 102363

Table 2 enzyme load for 2 h.

Identification and quantification of ions used in the GC/EI-MS-SIM determina­ The aLSLs were purified using molecular distillation device (KDL2,
tion of La and aLna. UIC GmbH, Alzenau, Germany), and the molecular distillation process
Window Time SIM ions (m/z) Quantity of Auxiliary was marginally modified from Zou et al. (2020). Distillation conditions
No. [min] ion (m/z) quantitative ion were as follows: feed temperature, 65 ◦ C; evaporation temperature,
(m/z) 185 ◦ C; condensation temperature, 50 ◦ C; scraper speed, 165 rpm; feed
1 3.0-6.0 129 157 200 157 129 rate, 4 mL/min; absolute pressure, 0.001 Pa. TLC was used to identify
2 6.0- 79 249 278 249 79 whether by-products (removal of fatty acids, monoglycerides and di­
15.0
glycerides) had been removed. The purified product was further held at
− 18 ◦ C for reserve.
quantitative determination of La and aLna was performed in selective
ion monitoring (SIM) mode. GC oven program (after 2 min at 150 ◦ C, the 2.5. Identification and analysis of aLSLs by UPLC-Q-TOF-MS
temperature was raised to 280 ◦ C at 15 ◦ C/min for 5 min), and SIM ions
of two substances were recorded at two time windows (Table 2) with a The purified aLSLs were dissolved with n-hexane (0.3 mg/mL).
solvent delay of 2 min. La and aLna were calculated from the standard UPLC-Q-TOF-MS (Waters, Milford, MA, USA) was used to determine
curve. TAG in aLSLs. The chromatography and mass spectrum conditions were
In this study, the La inserted content represents the content of La referred by Yu et al. (2022). TAG molecule species were identified by
embedded in the carbon backbone of FO triglycerides. The aLna cutting various quasi-molecular ions [M + NH4]+ and the corresponding
amount represents the amount of aLna being cleaved off the carbon [DAG]+ fragment from the MS/MS spectrum was selected. In this study,
backbone of FO triglycerides as free fatty acids. Formulas were calcu­ TAGs regioisomers weren’t distinguished. Fatty acids with a longer
lated as follows: carbon chain and higher saturation were ranked ahead.
La0 − La
La inserted content (%) = × 100 (2)
FO 2.6. Analysis of fatty acids (FAs) and Sn-2 position FAs

aLna cutting amount(%) =

aLna
× 100% (3) The FAs composition of the samples was transformed into fatty acid
FO methyl esters (FAME) according to Li et al. (2021), and analyzed by
GC-MS (Agilent ALS7693-GC7890B-MS7000D). The GC-MS operation
here, La0 was the mass of La in the system before the reaction, La was the
conditions were as follows: FAME samples were separated on an Agilent
mass of La in the system after the reaction, aLna was the mass of free
HP-5MS column (30.0 m × 0.25 mm × 0.25 μm) with He (purity
aLna in the reaction system, and FO was the mass of FO added to the
>99.999%) as carrier gas at a streaming rate of 1 mL/min. GC injector
system before the reaction.
temperature was 250 ◦ C, split flow ratio was 40:1, the initial tempera­
ture of the GC oven was 100 ◦ C and kept for 13 min, then ramped up to
2.4. Scale-up trial and purification of aLSLs 180 ◦ C at 10 ◦ C/min and maintained 6 min, then raised to 200 ◦ C at
1 ◦ C/min and held 20 min, and finally to 230 ◦ C at 4 ◦ C/min and
The scale-up trial was done at optimum parameters obtained from remained 10 min. The MS ionization voltage 70 eV, ion source tem­
verified RSM design. The aLSLs were synthesized by amplifying the perature 230 ◦ C, mass scanning range: 45–550 mass units (amu), solvent
amount of substrate by a factor of approximately 30 (about 630 mL). delay time 3 min. The proportion of each component in the sample was
Enzymatic reactions were run in a shaking water bath at 200 rpm/min obtained by the area normalization method and identified and queried
and 44.6 ◦ C, with 12 (La/FO) of substrate molar ratio and 9.11% of by the NIST 17 standard mass spectrometry library (Gao et al., 2017).

Fig. 1. Effects of different lipase on La inserted con­

tent and aLna cutting amount at different times. (a)
The La inserted content of 4 lipases at different times;
(b) The aLna cutting amount of 4 lipases at different
times; (c) The La inserted content of 4 lipases at 2 h;
(d) The aLna cutting amount of 4 lipases at 2 h ***,
represent extremely significant difference (P <
0.001), ** represent highly significant difference (P
< 0.01); *, represent significant difference (P < 0.05);
ns
, represent not significant (P > 0.05).

3
Y. Huang et al. Food Bioscience 52 (2023) 102363

Fig. 2. Effects of 2 h acidolysis reaction catalyzed by Lipozyme AOAB8 on La inserted content and aLna cutting amount under different conditions. (a) Effect of
reaction temperature on La inserted content and aLna cutting amount; (b) Effect of enzyme load on La inserted content and aLna cutting amount; (c) Effect of molar
ratio (La/FO) on the La inserted content and aLna cutting amount.

The components of Sn-2 FA were isolated as per Cai et al. (2015) and
Table 3
identified as described above.
BBD experimental design with the independent variables and experimental data
for the responses.
2.7. Fourier transform infrared (FTIR) analysis
Run synthesis Factors Experimental results

reaction molar enzyme La inserted aLna A Thermo Scientific®–Nicolet iS50 FTIR spectrophotometer (Wal­
temperature ratio (X2, load (X3, content (%) cutting
tham, Massachusetts, USA) was used to measure the FTIR spectrum in
(X1, ◦ C) La/FO) %) amount (%)
the infrared region (3500-500 cm− 1) of oil samples. It is equipped with a
1 60 10 6 14.24 50.17 diamond crystal cell attenuated total reflection (ATR) accessory and KBr
2 50 8 6 29.34 32.37
(potassium bromide) beam splitter. All the spectra were measured with
3 40 10 10 47.94 29.02
4 50 8 10 33.22 47.49 32 scans and 4 cm− 1 resolution.
5 50 10 8 56.5 21.11
6 50 10 8 55.25 25.41 2.8. Microstructure
7 60 12 8 15.97 30.09
8 40 12 8 46.27 8.68
9 50 12 10 48.37 10.59 All experiment samples were firstly melted at 70 ◦ C for 30 min to
10 50 12 6 37.14 31.04 remove the crystal memory, and then 10 μL melted oil was placed on a
11 40 10 6 20.46 44.46 preheated (about 70 ◦ C) glass slide and covered with a hot (about 70 ◦ C)
12 40 8 8 25.08 30.83
cover slide. Those preprocessed slices were immediately placed at
13 50 10 8 50.45 20.82
14 60 10 10 2.92 50.63
− 20 ◦ C for 24 h. All slices were captured by the microscopic imaging
15 50 10 8 47.02 21.47 system (Nikon eclipse Ci, Tokyo, Japan).
16 60 8 8 12.38 37.95
17 50 10 8 54.86 20.71
2.9. Thermal oxidation stability analysis

Thermogravimetric analysis (TG) was employed to evaluate the

oxidative thermal stability of the aLSLs. In brief, a simultaneous thermal

Table 4
ANOVA for response surface polynomial model of all independent variables.
Factor La inserted content aLna cutting amount

Sum of Squares df Mean Square F-Value p-value Sum of Squares df Mean Square F-Value p-value

Model 4734.24 9 526.03 51.16 <0.0001*** 2557.09 9 284.12 55.23 <0.0001***

Linear term
X1 1110.14 1 1110.14 107.97 <0.0001*** 390.09 1 390.09 75.83 <0.0001***
X2 284.89 1 284.89 27.71 0.0012** 581.88 1 581.88 113.11 <0.0001***
X3 122.34 1 122.34 11.9 0.0107* 51.51 1 51.51 10.01 0.0158*
Interactions
X1X2 77.35 1 77.35 7.52 0.0288* 51.02 1 51.02 9.92 0.0162*
X1X3 376.4 1 376.4 36.61 0.0005*** 63.13 1 63.13 12.27 0.01***
X2X3 13.51 1 13.51 1.31 0.2894ns 316.33 1 316.33 61.49 0.0001***
Quadratic
X21 1993.3 1 1993.3 193.86 <0.0001*** 347.93 1 347.93 67.63 <0.0001***
X22 158.28 1 158.28 15.39 0.0057** 71.12 1 71.12 13.82 0.0075**
X23 393.42 1 393.42 38.26 0.0005*** 665.8 1 665.8 129.42 <0.0001***
Residual 71.98 7 10.28 36.01 7 5.14
Lack of Fit 9.04 3 3.01 0.19 0.8971ns 20.28 3 6.76 1.72 0.3005ns
Pure Error 62.94 4 15.73 / / 15.73 4 3.93 / /
Cor Total 4806.22 16 / / / 2593.11 16 / / /
C.V. % 9.12 / / / / 7.52 / / / /
R-Squared 0.99 / / / / 0.99 / / / /
Adj R-Squared 0.97 / / / / 0.97 / / / /
Pred R-Squared 0.95 / / / / 0.87 / / / /
Adeq Precision 19.92 / / / / 25.40 / / / /

X1, X2, and X3 represent reaction temperature, the molar ratio of substrate, and the enzyme load; *** represent an extremely significant difference (P ≤ 0.001), **
represent a highly significant difference (P ≤ 0.01), * represent the significant difference (P ≤ 0.05), ns represents not significant (P ≥ 0.05).

4
Y. Huang et al. Food Bioscience 52 (2023) 102363

Fig. 3. Response surface plots of the combined effects of three independent variables (reaction conditions) on La inserted content (a, b, and c) and aLna cutting
amount (d, e, and f); Reaction conditions include reaction temperature (X1), the molar ratio of the substrate (X2) and the enzyme load (X3).

analyzer STA 449 F3 Jupiter (NETZSCH, Germany) was used to conduct 3. Results and discussion
and calibrated with indium. The oxidation stability of the sample (10.0
± 0.5 mg) was obtained under the following conditions: oxygen was 3.1. Effects of different reaction conditions on aLSLs synthesis
used as the low purging gas (at 20 mL/min), and the temperature was
dynamically raised from 0 ◦ C to 900 ◦ C at the rate of 10 ◦ C/min. Finally, 3.1.1. Lipase and time for the synthesis of aLSLs
the oxidation thermal stability curve was obtained. In this study, four commercial lipases, Lipozyme RM, Lipozyme TL
IM, Novozym 435 and Lipozyme AOAB8, representing enzymes with
2.10. Statistical analysis different regioselectivities and fatty acid preferences, were used as cat­
alysts. Among them, Novozym 435 is not site-selective while the rest
All reactions and tests were performed in triplicate. The measured lipases are region-specific. As acyl donor, La was embedded into the
data were analyzed using SPSS 22.0 software versions (SPSS Inc., Chi­ carbon backbone of FO triglyceride, and La inserted content and aLna
cago, IL, USA). Mean values were compared using Duncan’s multiple cutting amount were used as the reaction indexes to screen lipase and
range tests and multiple comparison tests (LSD). Data are expressed as reaction time (Fig. 1).
mean ± standard deviation, and P < 0.05 was considered a significant The inserted content of La and cutting amount of aLna (Fig. 1ab) was
difference. increased with the progress of the reaction and it reached the peak at
about 2–3 h (different enzymes were different). However, the reaction
efficiency starts to decrease after 3 h. At the initial stage of the reaction,
the reaction is mainly carried out in the positive direction due to the
high substrate concentration (Zou et al., 2020). Nevertheless, the
increased product content resulting in an increase in the reverse reaction
rate (Zhang et al., 2021). Considering to insert more La and cut less aLna

5
Y. Huang et al. Food Bioscience 52 (2023) 102363

Table 5 3.1.2. Effects of reaction temperature

Triacylglycerol (TAG) composition in aLSLs. The affinity between enzyme and substrate, the stability of the
TAG relative amount (%) TAG relative amount (%) enzyme, and the content of by-products are affected by reaction tem­
perature (Qin et al., 2012). The lower reaction temperature is a disad­
La-La-La 1.00 O-P-M 0.06
Ln-La-La 9.89 O-O-L 1.61 vantage to the improvement of enzyme catalytic ability, while a higher
Ln-Ln-La 6.11 O-L-P 0.94 temperature may increase energy consumption and generation of
L-La-La 5.65 S-O-Ln 2.05 by-products (Zou et al., 2020). Besides, the reaction temperature would
Ln-Ln-Ln 9.73 O-P-P 0.44 affect the viscosity of the reaction system, and then affect the interaction
L-Ln-La 4.56 O-O-P 0.93
O-La-La 4.87 O-O-O 0.94
between molecules and the efficiency of enzymatic reaction (Qin et al.,
L-Ln-Ln 7.10 S-L-P 0.09 2012). In this study, the effect of temperature on the synthesis of aLSLs
O-Ln-La 6.19 O-O-P 0.25 was investigated at 40–80 ◦ C. The results showed that with the increase
O-Ln-Ln 12.26 S-O-L 0.72 in temperature, La inserted content first increased and then decreased,
O-L-La 2.00 S-S-Ln 0.36
and the insertion amount were reached the maximum at 50 ◦ C (Fig. 2a).
O-L-Ln 7.29 S-P-P 0.32
S-Ln-Ln 2.26 S-O-P 1.07 Meanwhile, the aLna cutting amount was proportional to the increase in
O-O-La 1.17 S-O-O 1.21 temperature, and the aLna cutting amount was low at 40–60 ◦ C (Fig. 2a).
O-O-Ln 5.69 S-S-P 0.29 These results may be attributed to the higher molecular activity at high
S-L-Ln 1.86 S-S-O 0.69 temperature, which increased the solubility of the reactants and reduces
S-O-La 0.39 total 1.00
the viscosity of the mixture while ensuring enzyme activity (Dodson
sort out et al., 1992). However, high temperature also increases the changes in
GTL 1.00
the internal structure of triglycerides and accelerates the reaction rate of
MLCT 40.82
ɑ-MLCT 26.75 acid hydrolysis (Zou et al., 2014), thereby increasing the amount of the
ɑ-TAG 75.35 aLna cutting amount. In summary, in the synthesis of aLSLs, low tem­
S-TAG 1.62 perature (40–50 ◦ C) was beneficial. Considering the maximum yield of
US-TAG 98.38 aLSLs, 50 ◦ C was chosen for the following RSM study.
La, lauric acid (12:0); M, myristic acid (14:0); P, palmitic acid (16:0); S, stearic
acid (18:0); O, oleic acid (18:1); L, linoleic acid (18:2); Ln, α-linolenic acid 3.1.3. Effects of enzyme load
(18:3). GTL, triglyceride laurate; MLCT, medium- and long-chain tri­ A large number of enzyme loads could promote SLs synthesis by
acylglycerol; ɑ-MLCT, MLCT with aLna; ɑ-TAG, TAG with ɑ-linolenic acid; S- shortening the reaction time and weakening the influence of acyl
TAG, all three are saturated fatty acid TAG; US-TAG, TAG containing unsatu­ migration (Cai et al., 2015). As shown in Fig. 2b, La inserted content
rated FAs. continuously increased from 4% to 10% of the enzyme load, while it
began to decrease as the enzyme load continued to increase. The results
in aLSLs synthesis, the reaction time was determined to be 2 h. In the La were similar to those reported by Li et al. (2021). Within a reasonable
insertion (Fig. 1c), the activity of lipase AOAB8 was the highest, which range, the increase of enzyme loading will provide more reaction sites
significantly higher than that of Lipozyme TL IM and Novozym 435. and improve the reaction efficiency. Meanwhile, the aLna cutting
Moreover, the enzyme capacity to cut aLna (Fig. 1d) was opposite to the amount was decreased at first and then increased (Fig. 2b). Previous
La insertion. In all, considering the cost and catalytic efficiency, Lip­ studies also reported that the amount of lipase selected for C18 unsatu­
ozyme AOAB8 and 2 h of reaction time were selected for further study. rated FA in the acidolysis reaction was proportional to the enzyme load
(Karabulut et al., 2010). However, the excessive enzyme load leads to
biocatalyst agglomeration (Meng et al., 2013) and increases the cost,

Fig. 4. Upset plot of differences in TAG composition of aLSLs. The diagram shows the matrix layout of all the intersections of the 6 categories of TAG, with the black
circles in the matrix indicating that the set is part of the intersection. GTL, triglyceride laurate; MLCT, medium- and long-chain triacylglycerol; ɑ-MLCT, MLCT with
aLna; ɑ-TAG, TAG with ɑ-linolenic acid; S-TAG, all three are saturated fatty acid tags; US-TAG, TAG containing unsaturated FAs.

6
Y. Huang et al. Food Bioscience 52 (2023) 102363

Table 6 the peak when the molar ratio of La/FO was 8. Synchronously, with the
Change in total fatty acid composition and Sn-2 fatty acid distribution. increase of the molar ratio of La/FO, the amount of aLna cleaved into
FAs profiles Total FAs (% peak area) Sn-2 FAs (% peak free FA in FO triglyceride decreased first and then increased. When the
area) molar ratio of La/FO was 8, the amount of free aLna was the least.
aLSLs PM FO GTL aLSLs FO Similarly, Zou et al. (2020) prepared DHA-Rich MLCTs by
lipase-catalyzed acidolysis of microbial oil from Schizochytrium sp. with
Capric acid / / / 0.09 / /
(C10:0) ± 0.00
caprylic. Their results indicated that the higher substrate molar ratio
Undecanoic acid / / / 0.02 / / (caprylic/microbial oil) accelerated the initial reaction rate, but the
(C11:0) ± 0.00 excessive molar ratio resulted in the decrease of TAG content due to the
Dodecanoic acid 22.38 20.26 / 99.77 35.01 / limited number of reaction sites. In this study, the initial increase of La
(C12:0) ± 0.02 ± 0.02 ± 0.01 ± 0.76
inserted content might be related to the frequency of collisions between
Tridecanoic acid / / / 0.02 / /
(C13:0) ± 0.00 the substrates and enzyme heighten with the La/FO molar ratio
Myristic acid / / / 0.09 / / increasing. However, excessive La could not further promote the reac­
(C14:0) ± 0.01 tion when the equilibrium is reached. Therefore, the molar ratio of the
Palmitic acid 2.57 4.52 5.67 / 1.45 0.69 substrate, 8 of molar ratio (La/FO) was selected for the subsequent
(C16:0) ± 0.00 ± 0.01 ± 0.00 ± 0.02 ± 0.01
Stearic acid 2.07 3.57 4.04 / 1.12 0.57
study.
(C18:0) ± 0.00 ± 0.00 ± 0.01 ± 0.08 ± 0.06
Oleic acid 15.48 14.70 18.58 / / /
(C18:1n9c) ± 0.01 ± 0.02 ± 0.00 3.2. Optimization for aLSLs synthesis conditions
Linoleic acid 16.99 15.81 20.04 / 14.30 22.72
(C18:2n6c) ± 0.01 ± 0.02 ± 0.01 ± 0.28 ± 0.78 3.2.1. RSM model fitting
γ-linolenic acid 0.68 0.71 0.88 / / /
Based on the single-factor experiment, the limits and centroid values
(C18:3n6) ± 0.00 ± 0.01 ± 0.01
α-linolenic acid 39.78 40.02 50.15 / 48.13 76.02 of three independent variables were determined, and the lipase acid­
(C18:3n3) ± 0.03 ± 0.04 ± 0.01 ± 0.43 ± 0.71 olysis reaction was optimized by Box-Behnken design (BBD). The
Arachidic acid / 0.13 0.10 / / / experimental modeling results of La inserted content and aLna cutting
(C20:0) ± 0.00 ± 0.00
amount were shown in Table 3. The La inserted content and aLna cutting
Cis-11- 0.09 0.12 0.13 / / /
eicosenoic ± 0.00 ± 0.00 ± 0.00
amount ranged from 2.92% to 56.50% and 8.68%–50.63%, respectively.
acid (C20:1) As displayed in Table 4, the R2 values of La inserted content and aLna
Cis-11,14- / / 0.06 / / / cutting amount were 0.99 (P < 0.0001) and 0.99 (P < 0.0001),
eicosadienoic ± 0.00 respectively, which indicated that the model was extremely significant
acid (C20:2)
at a 95% confidence level. The lack of fit P values of the two models were
Octadecanoic / / 0.16 / / /
acid (C21:0) ± 0.00 0.90 and 0.30 respectively, which were insignificant at the α = 0.05
Elm acid (C22:0) / 0.11 0.11 / / / level. This implied that there is insignificant relationship between each
± 0.00 ± 0.00 pure error and its lack of fit. Additionally, the coefficient of variation
Xylic acid / / 0.08 / / /
(CV) of each model was quite low (CV < 10%), which represented the
(C24:0) ± 0.01
MCFA 22.38 20.27 / 99.88 35.08 /
good fit and reproducibility of the model (Liu et al., 2013). Meanwhile,
± ± ± ± 0.76 two models corrected correlation coefficients R2Adj and R2pred were closed
0.02b 0.02c 0.01a to 1, which indicated that uncertain factors had little influence on the
NM-SFA 4.65 8.32 10.16 0.11 2.57 1.26 experimental results and the fitting degree was good. Therefore, it could
± 0.0a
be considered that these two mathematical models were statistically
± ± ± ± ±
0.00c 0.01b 0.01a 0.00d 0.07b
MUFA 15.57 14.85 18.72 / / / acceptable, which could predict the internal relationship between
± ± ± acidolysis reaction process parameters and response value adequately.
0.01b 0.01c 0.01a Similarly, the RSM model was applied in other chemical reactions,
PUFA 57.45 56.53 71.13 / 62.43 98.74 Delgado Naranjo et al. (2021) optimized the enzymatic synthesis of an
± ± ± ± ±
0.04b 0.04c 0.01a 0.70b 0.07a
SLs rich in DHA at the sn-2 site from Aurantiochytrium limacinum SR21
La + aLna 62.16 60.27 50.15 99.77 83.14 76.02 oil and caprylic acid using the RSM model, which adequately described
± ± ± ± ± ± the reaction behavior (R2 > 0.91) and obtained SLs with caprylic acid
0.05b 0.04c 0.01d 0.01a 0.34a 0.71b incorporation of 40.8 mol% and DHA at the sn-2 position accounting for
FAs, Fatty acids; MCFA, medium chain FAs; NM-SFA, non-medium chain satu­ 45.5 mol%. Gokce et al. (2013) used the RSM to improve the acidolysis
rated F As; MUFA, monounsaturated FAs; PUFA, polyunsaturated FAs; La + reactions of Echium seed oil and La to prepare low-calorie SLs, the sta­
aLna, the total sum of La and aLna; aLSLs, synthesized medium and long chain tistical model of the R2 for La incorporation was 0.97 and the obtained
structural lipids of rich in aLna and La; PM, physical mixture oil, consisting of a SLs had a La content of 42.8%. In this study, the second-order poly­
simple physical mixture of FO and GTL, with a fatty acid ratio similar to aLSLs; nomial equations (4) and (5) described the relationship between reac­
FO, flaxseed oil; GTL, triglyceride laurate. Values (mean ± SD, n = 3); a, b, c, and tion temperature (X1), the molar ratio of the substrate (X2), and the
d
, small letters mean significant difference (P ≤ 0.05); /, not detected.
enzyme load (X3).

which were not conducive to the synthesis of aLSLs. Therefore, 10% La inserted content = 52.81 − 11.78X1 + 5.97X2 + 3.917X3 − 4.4X1 X2
enzyme load was chosen for the following RSM study. − 9.7X1 X3 + 1.84X2 X3 − 21.76X1 2 − 6.13X2 2 − 9.67X3 2 (4)

3.1.4. Effects of substrate molar ratio aLna cutting amount = 21.91 + 6.98X1 − 8.53X2 − 2.54X3 + 3.57X1 X2
The substrate molar ratio is an important variable in enzyme-
+ 3.97X1 X3 − 8.89X2 X3 + 9.09X1 2 − 4.11X2 2 + 12.57X3 2 (5)
catalyzed acidolysis reversible reactions, which affects the equilibrium
of the reaction (Lv et al., 2021). Especially, the increase in the amount of
3.2.2. Response surface optimization for synthesis conditions
the FA acyl donor is beneficial to its insertion into the triglyceride
From the regression evaluation analysis, the three independent
(Tecelão et al., 2019). The influence of the molar ratio of La/FO on the
variables have a linear effect on La inserted content and aLna cutting
synthesis of aLSLs was shown in Fig. 2c. The La inserted content reached
amount within the experimental range. As shown in Table 4, all linear

7
Y. Huang et al. Food Bioscience 52 (2023) 102363

Fig. 5. FTIR spectra of aLSLs, PM, FO, GTL and La at room temperature (25 ◦ C). aLSLs, synthesized medium and long chain structural lipids of rich in aLna and La;
PM, physical mixture oil, consisting of a simple physical mixture of FO and GTL, with a fatty acid ratio similar to aLSLs; FO, flaxseed oil; GTL, triglyceride laurate.

parameters (X1, X2, X3) and quadratic parameters (X12, X22, X32) were which 75.35% of TAGs contained aLna, and 40.82% of MLCTs contained
significant (P < 0.05 or P < 0.001). Except for X2X3 of La inserted La. However, Zou et al. (2020) also used UPLC-Q-TOF-MS to identify the
content interaction terms, all interaction terms were significant (P < TAG of 35 kinds of TAG with MCFA, accounting for 55.35%. It may be
0.05). In all, it was shown that the synthesis conditions (reaction tem­ related to the shorter reaction time of 2 h which is much less than their 6
perature, substrate molar ratio, and enzyme addition) were the main h. In this study, TAGs containing both aLna and La (ɑ-MLCT) accounted
factors affecting the synthesis of aLSLs. for 26.75% (65.53% of the total MLCTs), with Ln-La-La being the most
The three-dimensional (3D) response surface plots (Fig. 3) were abundant at 9.89%, meeting the medium- and long-chain fatty acid
constructed, which was convenient to analyze the relationship between edible oil and new resource food announcement standard. Therefore, La
independent variables and dependent variables (Sang et al., 2017). In and alna-rich MLCTs were successfully obtained in this study.
Fig. 3, the interaction of the two response values between the two var­
iables and their optimal ranges can be seen. La inserted content was
changed with the increase of the reaction factor and reached the highest 3.4. Composition and distribution of FAs in aLSLs
value near the midpoint of the response graph (Fig. 3a-c). The trend of
aLna cutting amount was opposite to La inserted content (Fig. 3d-f). The composition and distribution of FAs have a great impact on the
According to the purpose of our experiment, it was hoped that the least physical properties and oxidation stability of edible oil (Zhang et al.,
amount of aLna was cut into free FA in the SLs, while more lauric acid 2022). The composition and distribution of FAs in aLSLs, physical
can be inserted. The response values were simultaneously optimized by mixture oil (PM, consists of a simple physical mixture of flaxseed oil and
design-Expert-v8.0.6.1, and the optimal synthesis conditions were found triglyceride laurate, with a fatty acid ratio similar to aLSLs), FO, and GTL
as follows: reaction temperature 44.6 ◦ C, molar ratio 12 (La/FO), and (triglyceride laurate) were listed in Table 6. Twelve FAs were detected in
enzyme load 9.11%. It was expected to achieve 58.17% of La inserted FO used as raw material for aLSLs synthesis. Unsaturated fatty acids are
content and only 2.57% of aLna cutting amount. the main FAs, of which the most important FA aLna accounted for
50.15%, which was higher than that of vegetable oils such as walnut oil,
3.2.3. Verify the optimal reaction conditions corn oil, and rapeseed oil (Dubois et al., 2007). The sn-2 fatty acid in FO
To verify the reliability of the model, experiments were carried out was mainly polyunsaturated FAs (PUFA), in which aLna and linoleic
under the optimal conditions of prediction, and the results were 56.38 acid were 76.00% and 22.72% respectively. In the aLSLs, the main FAs
± 3.75% and 1.63 ± 0.26%, respectively. The experimental values were unsaturated, and the specific distribution was aLna (39.78%),
(58.17%, 2.57%) were within the 95% confidence interval of the pre­ linoleic acid (16.99%), and oleic acid (15.48%). This composition was
dicted values, which indicated that the strong correlation between the similar to FO, accounting for 73.022% of the total FAs. However, the
actual values and the predicted values, confirmed that the response saturated FAs in aLSLs were mainly La, which replaced stearic and
model was accurate and enough to reflect the expected optimization of palmitic acids in FO with a content of 22.38%. The content of aLna and
aLSLs synthesis. MCFA in aLSLs was higher than the SLs obtained by Zhao et al. (2013).
Briefly, they synthesized the SLs using camphor seed oil and perilla oil
under the catalysis of Lipozyme TL IM, and the aLna and MCFA contents
3.3. Composition of TAG in aLSLs in their SLs were 30.62% and 18.72%, respectively. Moreover, the sn-2
FAs in aLSLs were mainly composed of aLna (48.13%), La (35.01%), and
aLSLs were purified by molecular distillation, and the TAG compo­ linoleic acid (14.30%), and the content of PUFA was 62.43%. La was
sition of the product was analyzed by UPLC-Q-TOF-MS. The specific inserted, which was attributed to the application of the sn-1,3-specific
structural results and upset plot classification analysis were shown in Lipozyme AOAB8 lipase, which catalyzed the occurrence of acyl trans­
Table 5 and Fig. 4. A total of 33 types of TAGs were identified, among fer during transesterification (Zhang et al., 2022). In addition, it can be

8
Y. Huang et al. Food Bioscience 52 (2023) 102363

Fig. 6. Crystal morphology of aLSLs, PM, FO, GTL, and La at − 40 ◦ C for 24 h aLSLs, synthesized medium and long chain structural lipids of rich in aLna and La; PM,
physical mixture oil, consisting of a simple physical mixture of FO and GTL, with a fatty acid ratio similar to aLSLs; FO, flaxseed oil; GTL, triglyceride laurate.

seen from Table 6 that the composition and relative content of the main and PM had similar compositional ratios in FAs (Table 6), packing vi­
FAs in the PM were similar to those in aLSLs. The purity of GTL is higher brations and asymmetric stretching vibrations were stronger in aLSLs,
than 99.5%, and which FA was La. Therefore, the aLSLs obtained in our with significantly higher unsaturation and structural closer to FO.
study were rich in La + aLna, and the reaction model was an effective A pair of peaks near 2800-3000 cm− 1 represents the symmetrical and
synthesis method. asymmetric stretching vibrations of the CH bonds of CH2 aliphatic
groups. The frequency of the CH2 stretching band of the acyl chain de­
3.5. FTIR and microstructure analysis of aLSLs pends on the degree of conformational disorder, which is accompanied
by the increase of the infrared absorption band frequency caused by the
FTIR spectroscopy is used to monitor subtle changes in the structure symmetric and asymmetric stretching vibration of the lipid hydrocarbon
and function of liposomes (Mady et al., 2012). Because of the differences chain (Mady et al., 2012). Compared with PM, GTL, and La (2920 and
in the physicochemical properties and spatial location of fatty acids, 2850 cm− 1), it was observed that the corresponding peaks of aLSLs
lipids precise position and shape of FTIR peak are influenced (Rohman (2930 and 2860 cm− 1) shift to FO (2930 and 2860 cm− 1). Intuitively,
et al., 2014). The FTIR spectra of the 3500-500 cm− 1 mid-infrared region this happens should be due to the change of the type and position of FAs
of aLSLs, PM, FO, GTL, and La were shown in Fig. 5. Since almost all in the triglyceride structure by aLSLs synthesis, which leads to the for­
edible oils extracted from plants constitute triglycerides (Jamwal et al., mation of conformational chaos in the acyl chain, while simple physical
2021), the entire spectral range appears to be similar. It exhibited the mixing has only little effect on acyl sequencing and spatial
typical absorption band characteristics of the FTIR spectrum of common conformation.
edible oils and fats, with two main “information” regions at 3100-2800 The absorption band of ester C–– O is sensitive to the polarity change
cm− 1 and 1800-500 cm− 1. However, the absorption peaks of 3010 cm− 1 of its local environment and is affected by hydrogen bonds and other
(attributed to the symmetric stretching vibration of cis-olefinic groups, interactions (Mady et al., 2012). Compared with the vibration frequency
=CH) and 1660 cm− 1 (the stretching vibration of disubstituted cis C– –C of 1740 cm− 1 (stretching vibration of C–– O group) of PM and GTL, the
assigned to unsaturated acyl groups) were observed in aLSLs, PM and FO aLSLs directly shifted to 1750 cm− 1 of FO, indicating that the insertion
stretching vibration), which were not observed in GTL and La, indicated of La in the glycerol molecule of FO didn’t change the polarity of the
that the presence of unsaturated fatty acids in aLSLs. But now, the peak aLSLs. It may be that the long-chain fatty acid acyl groups in aLSLs bury
intensities in PM were distinctly smaller, indicating that although aLSLs the effect of La acyl groups on molecular polarity, while physical mixing

9
Y. Huang et al. Food Bioscience 52 (2023) 102363

Fig. 7. TG-DTG curve of aLSLs, PM, FO, and GTL under gas flow at a heating rate of 10 ◦ C/min aLSLs, synthesized medium and long chain structural lipids of rich in
aLna and La; PM, physical mixture oil, consisting of a simple physical mixture of FO and GTL, with a fatty acid ratio similar to aLSLs; FO, flaxseed oil; GTL, tri­
glyceride laurate.

Table 7
Results of dynamic Thermogravimetric (TG) measurements for aLSLs, PM, FO and GTL.
Parameter aLSLs PM FO GTL

Mass Increase Onset (◦ C) 165.47 ± 2.42b 159.23 ± 3.58bc 158.57 ± 3.12c 183.77 ± 4.13a
Peak Temp. (◦ C) 192.47 ± 1.8c 200.93 ± 3.32ab 198.37 ± 4.95bc 205.43 ± 1.99a
Peak weight increase (%) 101.18 ± 0.12b 101.02 ± 0.19b 101.74 ± 0.44a 100.29 ± 0.17c
Mass Loss Mass Loss Onset (◦ C) 281.67 ± 11.34b 263.03 ± 9.04c 359.67 ± 5.09a 275.30 ± 9.65bc
Mid (◦ C) 386.1 ± 3.52b 351.87 ± 14.5c 425.9 ± 6.82a 303.73 ± 4.73d
Inflection (◦ C) 416.3 ± 6.02a 318.43 ± 24.67b 427.27 ± 0.45a 311.7 ± 7.27b
End (◦ C) 494.3 ± 8.64a 438.27 ± 13.59b 493.83 ± 19.03a 336.67 ± 16.4c

aLSLs, synthesized medium and long chain structural lipids of rich in aLna and La; PM, physical mixture oil, consisting of a simple physical mixture of FO and GTL, with
a fatty acid ratio similar to aLSLs; FO, flaxseed oil; GTL, triglyceride laurate; Values (mean ± SD, n = 3); a, b, c and d, small letters mean significant difference (P ≤ 0.05).

does not involve changes in molecular structure and preserves the po­ et al. (2018), where the crystal structure differed significantly between
larity of original La acyl groups. Moreover, La had a distinct charac­ the physical mixture oil and the structural lipids. Meanwhile, the crystal
teristic vibration frequency of carboxylic acid carbonyl group 1700 microstructures of La were completely different from others. The
cm− 1 (Lazzari & Chiantore, 1999). In conclusion, aLSLs were more experimental results showed that the microstructure of oils was related
similar to FO in FTIR spectra, while PM was similar to GTL, both were to their molecular structure. Through the above FTIR analysis, know
different from La. It can be concluded that the structure, polarity, and that the structure, polarity, and properties of aLSLs are close to FO.
properties of aLSLs were close to FO, while PM changes the composition Although PM changes the composition ratio of FAs similar to aLSLs, it
ratio of FAs to be similar to aLSLs, but has less effect on the property, had little effect on the structure and molecular polarity. Similar results
structure and, molecular polarity. had been obtained here. All samples had different crystal structures, but
The microstructure of fat is directly determined by FAs and TAG in the microstructure of the aLSLs was similar to the FO and quite distinct
the crystal network, which has different effects on the physical proper­ from the PM and GML, PM was similar to the GML. This might also be
ties such as melting point, solubility, and chemical potential, and play an the reason for their different stability.
important role in the macroscopic properties of fat (Narine & Hum­
phrey, 2004). The microstructures and crystal morphologies of the
crystal networks was shown in Fig. 6. It was observed that aLSLs and FO 3.6. Study on thermal oxidation stability
have relatively similar crystal morphology, and the overall structure was
a dispersed rod structure. Due to the more and more complex molecular During storage or heat treatment, oils and fats are prone to oxidation
structures and types of aLSLs, there were more types of crystals of and thermal decomposition. Oxidation products may have adverse ef­
different sizes. The crystal morphology of PM and GTL were very similar, fects on the human body, while thermal decomposition may release
and the tiny needle-like aggregates formed snowflake spherical crystals fumes containing toxic and harmful free radicals. From the point of view
with different densities, among which the density of the crystal clusters of safety, the oxidation stability of the oil is one of the most important
of PM was lower. Nevertheless, there were distinct differences in the parameters (Symoniuk et al., 2016). Thermogravimetric (TG) method is
crystal morphology of aLSLs and PM in the presence of similar fatty acid used to study the thermal decomposition of rancidity oil (Rudnik et al.,
ratios (Table 6), and similar results were obtained in the article by Xu 2001). The dynamic TG-DTG curve was shown in Fig. 7. From the TG
curve, it could be observed that the mass of aLSLs, PM, and FO increases

10
Y. Huang et al. Food Bioscience 52 (2023) 102363

slightly at first from the temperature range of 150–210 ◦ C, and then Declaration of competing interest
decreases rapidly, ending at the temperature range of 450–500 ◦ C, the
mass loss reaches 95–98%. The onset temperature (TOnset, TG) at which The authors declare no competing interests.
weight gain begins is the onset of oxidation (Rudnik et al., 2001),
whereas the onset temperature of weight loss (TOnset, Mass) is the onset of Data availability
thermal decomposition (Lazzari & Chiantore, 1999). The specific results
of thermo-oxidative stability were shown in Table 7. The TOnset, TG Data will be made available on request.
(165.47 ◦ C) of aLSLs was significantly higher in FO (158.57 ◦ C) and
slightly higher in PM (159.23 ◦ C), showing that aLSLs had a good pro­ Acknowledgment
tective effect on easily oxidized unsaturated fatty acids through struc­
turing, even with the same proportion of FA was also slightly better than The authors gratefully acknowledge the funding support of Key
simple physical mixing method. Simultaneously, the structuring of Project of Science and Technology Innovation 2025 program
aLSLs also helps to significantly enhance the ability to resist thermal (2019B10060), and the National Natural Science Foundation of China
decomposition, the TOnset, Mass (281.67 ◦ C) of aLSLs was significantly (32072224).
larger than PM (263.03 ◦ C) and slightly larger than GTL (275.3 ◦ C).
These results might since La is inserted into the TAG molecule during the References
structural synthesis of aLSLs, which has a certain protective effect on the
oxidative thermal stability of the unsaturated fatty acids in TAG Azizi, M., Kierulf, A., Lee, M. C., & Abbaspourrad, A. (2018). Improvement of
physicochemical properties of encapsulated echium oil using nanostructured lipid
(Meireles et al., 2022), which is better than simple physical mixing. carriers. Food Chemistry, 246, 448–456.
Similar results were obtained by Negishi et al. (2003), MLCT has better Bornscheuer, U. T. (2018). Enzymes in lipid modification. Annual Review of Food Science
cooking performance with higher smoke point and less tendency to foam and Technology, 9, 85–103.
Cai, H., Li, Y., Zhao, M., Fu, G., Lai, J., & Feng, F. (2015). Immobilization, regiospecificity
than simple physical mixtures. Furthermore, similar results were ob­ characterization and application of Aspergillus oryzae lipase in the enzymatic
tained for parameters in both the mass increase and the mass loss phase. synthesis of the structured lipid 1,3-dioleoyl-2-palmitoylglycerol. PLoS One, 10(7),
From the DTG curve (Fig. 7), it was observed that the thermos-oxygen Article e0133857.
Chang, M., Yang, J., Guo, X., Zhang, T., Liu, R., Jin, Q., & Wang, X. (2021). Medium/
decomposition of aLSLs and FO has three decomposition stages, while long-chain structured triglycerides are superior to physical mixtures triglycerides in
PM shows four, and GTL has only one. The different stages of Caenorhabditis elegans lifespan through an AMPK modified pathway. Food Bioscience,
thermo-oxidative decomposition of these oils are related to the type, 39, Article 100815.
Delgado Naranjo, J. M., Jiménez Callejón, M. J., Peñuela Vásquez, M., Rios, L. A., &
unsaturation, and structure of FAs contained (Sajeeb & Rajendrakumar,
Robles Medina, A. (2021). Optimization of the enzymatic synthesis of structured
2019). Therefore, the different thermal decomposition phenomena of triacylglycerols rich in docosahexaenoic acid at sn-2 position by acidolysis of
aLSLs and PM may be because La in SLs is mainly inserted into TAG Aurantiochytrium limacinum SR21 oil and caprylic acid using response surface
containing long-chain FA, while PM maintains the original TAG struc­ methodology. Journal of Applied Phycology, 33(4), 2031–2045.
Dodson, G. G., Lawson, D. M., & Winkler, F. K. (1992). Structural and evolutionary
ture and also has its original thermal oxygen decomposition behavior. In relationships in lipase mechanism and activation. Faraday Discussions, 93, 95–105.
all, the thermal oxygen decomposition behavior of aLSLs was signifi­ Dubois, V., Breton, S., Linder, M., Fanni, J., & Parmentier, M. (2007). Fatty acid profiles
cantly improved by altering the structure of TAG through acidolysis of 80 vegetable oils with regard to their nutritional potential. European Journal of
Lipid Science and Technology, 109(7), 710–732.
synthesis reaction. Farag, M. A., Elimam, D. M., & Afifi, S. M. (2021). Outgoing and potential trends of the
omega-3 rich linseed oil quality characteristics and rancidity management: A
4. Conclusion comprehensive review for maximizing its food and nutraceutical applications. Trends
in Food Science & Technology, 114, 292–309.
Gao, X., Lv, S., Wu, Y., Li, J., Zhang, W., Meng, W., Wang, C., & Meng, Q. (2017). Volatile
In this study, La was embedded into FO to synthesize mid- and long- components of essential oils extracted from Pu-erh ripe tea by different extraction
chain structural esters rich in aLna and La through transesterification by methods. International Journal of Food Properties, 20, S240–S253.
Gokce, J., Yesilcubuk, N. S., & Ustun, G. (2013). Enzymatic production of low-calorie
acidolysis. The optimal parameters were optimized by single factor ex­ structured lipid from Echium seed oil and lauric acid: Optimisation by response
periments and the RSM method Box-Behnken design. Under the condi­ surface methodology. International Journal of Food Science and Technology, 48(7),
tion of using Lipozyme AOAB8 enzyme reaction for 2.0 h, the best 1383–1389.
Guo, Y., Cai, Z., Xie, Y., Ma, A., Zhang, H., Rao, P., & Wang, Q. (2020). Synthesis,
parameter combination is reaction temperature 44.6 ◦ C, substrate molar
physicochemical properties, and health aspects of structured lipids: A review.
ratio 12 (La/FO), and enzyme load 9.11%. The aLSLs were purified by Comprehensive Reviews in Food Science and Food Safety, 19(2), 759–800.
molecular distillation, and 13 types of TAGs containing aLna with a Guo, X., Zhang, T., Shi, L., Gong, M., Jin, J., Zhang, Y., Liu, R., Chang, M., Jin, Q., &
content of 75.35%, and 9 types of MLCTs containing La with a content of Wang, X. (2018). The relationship between lipid phytochemicals, obesity and its
related chronic diseases. Food & Function, 9(12), 6048–6062.
40.82%. In addition, the La and aLna in aLSLs were 22.38% and 39.78%, Jamwal, R., Kumari, S., Kelly, S., Cannavan, A., & Singh, D. K. (2021). Non-targeted
respectively, and the physicochemical properties and microstructure of fingerprinting approach for rapid quantification of mustard oil adulteration with
aLSLs were similar to FO. The thermo-oxidative decomposition behavior linseed oil: An economically motivated adulteration. Vibrational Spectroscopy, 113,
Article 103226.
of SLs was significantly improved due to the insertion of La. The cor­ Ji, S., Xu, F., Zhang, N., Wu, Y., Ju, X., & Wang, L. (2021). Dietary a novel structured
relation between FAs and thermal-oxygen decomposition was analyzed, lipid synthesized by soybean oil and coconut oil alter fatty acid metabolism in
and the improvement of thermal-oxygen stability was related to the C57BL/6J mice. Food Bioscience, 44, Article 101396.
Karabulut, I., Durmaz, G., & Hayaloglu, A. A. (2010). C18 unsaturated fatty acid
insertion of La and the structure of TAG, and also provided reference selectivity of lipases during the acidolysis reaction between tripalmitin and oleic,
value for the rational application of aLna-enriched SLs. At the same time, linoleic, and linolenic acids. Journal of the American Oil Chemists’ Society, 87(11),
it provides a reference value for the rational application of rich aLSLs. 1301–1307.
Lazzari, M., & Chiantore, O. (1999). Drying and oxidative degradation of linseed oil.
Polymer Degradation and Stability, 65(2), 303–313.
Credit authorship contribution statement Lee, Y., Tang, T., Chan, E., Phuah, E., Lai, O., Tan, C., Wang, Y., Ab Karim, N. A., Mat
Dian, N. H., & Tan, J. S. (2022). Medium chain triglyceride and medium-and long
chain triglyceride: Metabolism, production, health impacts and its applications–a
Ying Huang: Conceptualization, Methodology, Writing - original
review. Critical Reviews in Food Science and Nutrition, 62(15), 4169–4185.
draft, Software, Writing - review & editing. Wangxin Liu: Investigation, Li, Y., Li, C., Feng, F., Wei, W., & Zhang, H. (2021). Synthesis of medium and long-chain
Writing - review & editing. Xianliang Luo: Visualization, Investigation. triacylglycerols by enzymatic acidolysis of algal oil and lauric acid. LWT–Food
Minjie Zhao: Software, Data curation. Tao Liu: Software, Data curation. Science and Technology, 136, Article 110309.
Liu, Z., Dang, J., Wang, Q., Yu, M., Jiang, L., Mei, L., Shao, Y., & Tao, Y. (2013).
Fengqin Feng: Project administration, Resources, Writing - review & Optimization of polysaccharides from Lycium ruthenicum fruit using RSM and its anti-
editing. oxidant activity. International Journal of Biological Macromolecules, 61, 127–134.

11
Y. Huang et al. Food Bioscience 52 (2023) 102363

Luo, X., Liu, W., Zhao, M., Huang, Y., & Feng, F. (2022). Glycerol monolaurate beyond an Candida parapsilosis lipase/acyltransferase. Journal of the American Oil Chemists’
emulsifier: Synthesis, in vivo fate, food quality benefits and health efficacies. Trends Society, 96(7), 777–787.
in Food Science & Technology, 127, 291–302. Terada, S., Sekine, S., & Aoyama, T. (2015). Dietary intake of medium- and long-chain
Lv, W., Wu, C., Lin, S., Wang, X., & Wang, Y. (2021). Integrated utilization strategy for triacylglycerols prevents the progression of hyperglycemia in diabetic ob/ob mice.
soybean oil deodorizer distillate: Synergically synthesizing biodiesel and recovering Journal of Oleo Science, 64(6), 683–688.
bioactive compounds by a combined enzymatic process and molecular distillation. Terada, S., Yamamoto, S., Sekine, S., & Aoyama, T. (2012). Dietary intake of medium-
ACS Omega, 6(13), 9141–9152. and long-chain triacylglycerols ameliorates insulin resistance in rats fed a high-fat
Mady, M. M., Fathy, M. M., Youssef, T., & Khalil, W. M. (2012). Biophysical diet. Nutrition, 28(1), 92–97.
characterization of gold nanoparticles-loaded liposomes. Physica Medica, 28(4), Vors, C., Le Barz, M., Bourlieu, C., & Michalski, M. C. (2020). Dietary lipids and
288–295. cardiometabolic health: A new vision of structure–activity relationship. Current
McCarty, M. F., & DiNicolantonio, J. J. (2016). Lauric acid-rich medium-chain Opinion in Clinical Nutrition and Metabolic Care, 23(6), 451–459.
triglycerides can substitute for other oils in cooking applications and may have Wang, X., Qiu, S., Wang, X., Amuti, A., Zan, M., Li, Y., Wang, Z., & Dang, L. (2021).
limited pathogenicity. Open heart, 3(2), Article e000467. Developing peony (Paeonia suffruticosa Andr.) seed oil products based on α-linolenic
Meireles, B. R. L.d. A., Alcântara, M. A., Polari, I.d. L. B., Souza, A. G.d., Santos, N. A.d., acid by microemulsification. Industrial Crops and Products, 161, Article 113211.
Grisi, C. V. B., & Cordeiro, A. M. T.d. M. (2022). Catole coconut (Syagrus cearensis) Wang, Y., Zhang, T., Liu, R., Chang, M., Wei, W., Jin, Q., & Wang, X. (2022). Reviews of
oil: Physicochemical characterization and thermo-oxidative stability by TG/DTG/ medium- and long-chain triglyceride with respect to nutritional benefits and
DTA and rancimat. Journal of Thermal Analysis and Calorimetry, 147(5), 3591–3598. digestion and absorption behavior. Food Research International, 155, Article 111058.
Meng, Z., Geng, W., Li, J., Yang, Z., Jiang, J., Wang, X., & Liu, Y. (2013). Enzymatically Xu, Y., Zhu, X., Ma, X., Xiong, H., Zeng, Z., Peng, H., & Hu, J. (2018). Enzymatic
catalyzed synthesis of anti-blooming agent 1, 3-dibehenoyl-2-oleoyl glycerol in a production of trans-free shortening from coix seed oil, fully hydrogenated palm oil
solvent-free system: Optimization by response surface methodology. Journal of and Cinnamomum camphora seed oil. Food Bioscience, 22, 1–8.
Agricultural and Food Chemistry, 61(45), 10798–10806. Yang, J., Wen, C., Duan, Y., Deng, Q., Peng, D., Zhang, H., & Ma, H. (2021). The
Narine, S., & Humphrey, K. (2004). A comparison of lipid shortening functionality as a composition, extraction, analysis, bioactivities, bioavailability and applications in
function of molecular ensemble and shear: Microstructure, polymorphism, solid fat food system of flaxseed (Linum usitatissimum L.) oil: A review. Trends in Food Science
content and texture. Food Research International, 37(1), 28–38. & Technology, 118, 252–260.
Negishi, S., Itakura, M., Arimoto, S., Nagasawa, T., & Tsuchiya, K. (2003). Measurement Yue, C., Li, M., Li, J., Han, X., Zhu, H., Yu, G., & Cheng, J. (2020). Medium-, long- and
of foaming of frying oil and effect of the composition of TG on foaming. Journal of the medium-chain-type structured lipids ameliorate high-fat diet-induced
American Oil Chemists’ Society, 80(5), 471–474. atherosclerosis by regulating inflammation, adipogenesis, and gut microbiota in
Panpipat, W., Chaijan, M., & Guo, Z. (2018). Oxidative stability of margarine enriched ApoE(-/-) mice. Food & Function, 11(6), 5142–5155.
with different structures of β-sitosteryl esters during storage. Food Bioscience, 22, Yu, J., Wei, W., Wang, F., Yu, R., Jin, Q., & Wang, X. (2022). Evaluation of total, sn-2
78–84. fatty acid, and triacylglycerol composition in commercial infant formulas on the
Punia, S., Sandhu, K. S., Siroha, A. K., & Dhull, S. B. (2019). Omega 3-metabolism, Chinese market: A comparative study of preterm and term formulas. Food Chemistry,
absorption, bioavailability and health benefits–A review. Pharma Nutrition, 10, 384, Article 132477.
Article 100162. Zhang, Z., Lee, W. J., Xie, X., Ye, J., Tan, C. P., Lai, O. M., Li, A., & Wang, Y. (2021).
Qin, X., Yang, B., Huang, H., & Wang, Y. (2012). Lipase-catalyzed incorporation of Enzymatic interesterification of palm stearin and palm olein blend catalyzed by sn-1,
different fatty acids into tripalmitin-enriched triacylglycerols: Effect of reaction 3-specific lipase: Interesterification degree, acyl migration, and physical properties.
parameters. Journal of Agricultural and Food Chemistry, 60(9), 2377–2384. Journal of Agricultural and Food Chemistry, 69(32), 9056–9066.
Rohman, A., Riyanto, S., Sasi, A. M., & Yusof, F. M. (2014). The use of FTIR spectroscopy Zhang, Y., Liu, Y., Wang, J., Zhang, R., Jing, H., Yu, X., Zhang, Y., Xu, Q., Zhang, J., &
in combination with chemometrics for the authentication of red fruit (Pandanus Zheng, Z. (2010). Medium- and long-chain triacylglycerols reduce body fat and
conoideus Lam) oil from sunflower and palm oils. Food Bioscience, 7, 64–70. blood triacylglycerols in hypertriacylglycerolemic, overweight but not obese,
Rudnik, E., Szczucinska, A., Gwardiak, H., Szulc, A., & Winiarska, A. (2001). Chinese individuals. Lipids, 45(6), 501–510.
Comparative studies of oxidative stability of linseed oil. Thermochimica Acta, 370 Zhang, Z., Xie, X., Lee, W. J., Zhao, G., Li, C., & Wang, Y. (2022). The effects of
(1–2), 135–140. interesterification on the physicochemical properties of Pangasius bocourti oil and its
Sajeeb, A., & Rajendrakumar, P. K. (2019). Comparative evaluation of lubricant fractions. Food Chemistry, 371, Article 131177.
properties of biodegradable blend of coconut and mustard oil. Journal of Cleaner Zhao, M., Tang, L., Zhu, X., Hu, J., Li, H., Luo, L., Lei, L., & Deng, Z. (2013). Enzymatic
Production, 240, Article 118255. production of zero-trans plastic fat rich in α-linolenic acid and medium-chain fatty
Sang, J., Sang, J., Ma, Q., Hou, X., & Li, C. (2017). Extraction optimization and acids from highly hydrogenated soybean oil, cinnamomum camphora seed oil, and
identification of anthocyanins from Nitraria tangutorun Bobr. seed meal and perilla oil by lipozyme TL IM. Journal of Agricultural and Food Chemistry, 61(6),
establishment of a green analytical method of anthocyanins. Food Chemistry, 218, 1189–1195.
386–395. Zou, X., Hu, J., Zhao, M., Zhu, X., Li, H., Liu, X., Liu, R., & Deng, Z. (2014). Lipozyme RM
Sneha, R., & Jeyarani, T. (2018). Lipase-catalysed acidolysis of mango kernel fat with IM-catalyzed acidolysis of Cinnamomum camphora seed oil with oleic acid to produce
capric acid to obtain medium- and long-chain triacylglycerols. International Journal human milk fat substitutes enriched in medium-chain fatty acids. Journal of
of Food Science and Technology, 53(6), 1527–1534. Agricultural and Food Chemistry, 62(43), 10594–10603.
Symoniuk, E., Ratusz, K., & Krygier, K. (2016). Comparison of the oxidative stability of Zou, X., Ye, L., He, X., Wu, S., Zhang, H., & Jin, Q. (2020). Preparation of DHA-rich
linseed (Linum usitatissimum L.) oil by pressure differential scanning calorimetry and medium- and long-chain triacylglycerols by lipase-catalyzed acidolysis of microbial
Rancimat measurements. Journal of Food Science & Technology, 53(11), 3986–3995. oil from Schizochytrium sp. with medium-chain fatty acids. Applied Biochemistry and
Tecelão, C., Perrier, V., Dubreucq, E., & Ferreira-Dias, S. (2019). Production of human Biotechnology, 191(3), 1294–1314.
milk fat substitutes by interesterification of tripalmitin with ethyl oleate catalyzed by

12